Character Rankings List - Post-March 2010

[axiomnightmare]

I think that you should place weights on the tournaments. A huge tournament like XESTICLE should count for more than a small weekly.

 

[keeper]

Hooray for me helping Olimar?

 

[JJ259]

The main (only?) problem that I see with this list is the fact that most characters with high scores are just plain popular. If only there was a way to count in the number of x character when giving up the points.

Don't take this as an insult though, you're doing the best job with what you've been presented. Keep it up.

 

[Zankoku]

The main (only?) problem that I see with this list is the fact that most characters with high scores are just plain popular. If only there was a way to count in the number of x character when giving up the points.

Don't take this as an insult though, you're doing the best job with what you've been presented. Keep it up.

I suppose you have a good point. I altered the scoring a little to put greater emphasis on winning and top4 places.

 

[ChokE]

Crap Pit is C ranked. Right now we are seriously lagging pro players to represent Pit. I can’t believe is beaten by half of characters in SSBB.

 

[JJ259]

Double post x.x

 

[JJ259]

In the never ending quest to make my characters seem better than they are, I noticed mis calculations in both of them. if top 8 is worth 1, top4 is worth 3 and a win is worth 6.

Actually quite a few were off.

S Rank
Meta Knight (19 top8, 19 top4, 5 wins, 101 pts) 106
Snake (12 top8, 14 top4, 9 wins, 99 pts) 108
A Rank
Mr. Game & Watch (11 top8, 6 top4, 4 wins, 49 pts) 53
King Dedede (3 top8, 10 top4, 3 wins, 48 pts) 51
Wario (1 top8, 4 top4, 6 wins, 43 pts) 49
ROB (8 top8, 10 top4, 1 win, 43 pts) 44
Pikachu (3 top8, 8 top4, 2 wins, 37 pts) 39
Marth (10 top8, 5 top4, 2 wins, 35 pts) 37
B Rank
Wolf (7 top8, 7 top4, 28 pts) 28
Ike (7 top8, 5 top4, 1 win, 27 pts) 28
Kirby (7 top4, 1 win, 26 pts) 27
Toon Link (3 top8, 4 top4, 2 win, 25 pts) 27
Peach (2 top8, 6 top4, 1 win, 25 pts) 26
Fox (5 top8, 5 top4, 1 win, 25 pts) 26
Lucario (2 top8, 4 top4, 2 wins, 24 pts) 26
Olimar (8 top8, 5 top4, 23 pts) 23
Sonic (3 top8, 2 top4, 2 wins, 19 pts) 21
Falco (5 top8, 3 top4, 1 win, 19 pts) 20
Donkey Kong (5 top8, 1 top4, 2 win, 18 pts) 20
C Rank
Pit (4 top8, 2 top4, 1 win, 15 pts) 16
Ice Climbers (3 top8, 4 top4, 15 pts) 15
Lucas (4 top8, 3 top4, 13 pts) 13
Diddy Kong (4 top8, 3 top4, 13 pts) 13
Bowser (1 top8, 3 top4, 10 pts) 10
Luigi (2 top4, 6 pts) 6
Captain Falcon (2 top4, 6 pts) 6
Zero Suit Samus (3 top8, 1 top4, 6 pts) 6
Jigglypuff (3 top8, 1 top4, 6 pts) 6
Ness (2 top8, 1 top4, 5 pts) 5
Mario (2 top8, 1 top4, 5 pts) 5
D Rank
Sheik (1 top8, 1 top4, 4 pts) 4
Link (1 top8, 1 top4, 4 pts) 4
Zelda (1 top4, 3 pts) 3
Yoshi (1 top4, 3 pts) 3
Pokémon Trainer (1 top4, 3 pts) 3
Ganondorf (3 top8, 3 pts) 3
Samus (2 top8, 2 pts) 2

It's late and I'm bored...so i fix'd em =)
Also, apparently snake > mk

 

[Zankoku]

Sorry, it's supposed to say +5.

 

[Emblem Lord]

Hmmm, right now this does seem like more of a popularity contest. Wario is good, but he isn't THAT good. I feel the same way about Pikachu. And Pit can't be that low. There is no way.

But in like 2 more months or so this should be accurate so keep at it.

 

[Meta Dude]

uh, the data is most definitely biased. i can almost guarantee you that there are tons more mk/snake/marth players than jigglypuff/ness/etc. that should be a footnote.

 

[HiddenBowser]

The reason there is more is because the character is better.

 

[Meta Dude]

The reason there is more is because the character is better.

grats on missing the point. regardless of who (you think) is better, the data is still biased, for better or worse. there's already like five threads where you can discuss who's better and who's not. i was under the assumption that this thread was for compiling data on each character's performance at tournaments, in which case the more popular characters are always gonna be on top (probability wise), making it a pretty biased study.

i don't really have a problem at all with this thread, but it should at least be noted that there are much more people playing those top characters. saying "well they're better" doesn't make this any less biased.

 

[Zankoku]

grats on missing the point. regardless of who (you think) is better, the data is still biased, for better or worse. there's already like five threads where you can discuss who's better and who's not. i was under the assumption that this thread was for compiling data on each character's performance at tournaments, in which case the more popular characters are always gonna be on top (probability wise), making it a pretty biased study.

i don't really have a problem at all with this thread, but it should at least be noted that there are much more people playing those top characters. saying "well they're better" doesn't make this any less biased.

Obviously, if a character is seen more often in a tournament, they're much more relevant to the metagame. This is regardless of how good you might think Ness or Yoshi might be or eventually become. The list is now putting heavy emphasis on top4 and winning positions over top8 positions; I'm pretty sure the reason characters like Meta Knight and Snake are so popular is because they happen to be winning most of the tournaments.

Bias nothing, if a character is proven to be good, it's practically a guarantee that good players will pick that character up and take it somewhere near the top. It will take time, but you can't defend Ness' lackluster placings just by saying not many people play him - clearly, people do play him, but nobody's actually gotten farther than top4 with him.

 

[Meta Dude]

Obviously, if a character is seen more often in a tournament, they're much more relevant to the metagame. This is regardless of how good you might think Ness or Yoshi might be or eventually become. The list is now putting heavy emphasis on top4 and winning positions over top8 positions; I'm pretty sure the reason characters like Meta Knight and Snake are so popular is because they happen to be winning most of the tournaments.

Bias nothing, if a character is proven to be good, it's practically a guarantee that good players will pick that character up and take it somewhere near the top. It will take time, but you can't defend Ness' lackluster placings just by saying not many people play him - clearly, people do play him, but nobody's actually gotten farther than top4 with him.

first of all, i'm not defending ness' placing or saying MK/snake aren't good characters. all i pointed out was that this is biased, whether you like to admit it or not. by bias, i mean it's biased statistically, not by the author of the thread. if you've taken a stats class you know what i'm talking about.

this is just another thread for people obsessed with tier lists.

 

[Zankoku]

Well, clearly there's going to be bias. There's 35+ characters in the game, and they're not all going to arrange themselves into a perfectly normal distribution.

 

[Meta Dude]

exactly, there's always gonna be characters who are overrepresented, therefore making it more likely that they'll be on top. perhaps those characters are better and that's why they're overrepresented, but that will never change the fact that something like this will favor them. that's all i'm pointing out, i wasn't looking to debate ganondorf being top tier.

 

[JJ259]

The only way to really eliminate the bias would be to take the entire tournament roster and then divide the amount of points gotten by the top 8/4 and winners by the amount of that character in the entire roster...which is quite a bit too much to ask. In my opinion, if you really want to see how good a specific character does, don't only look at the total points but look at what % they are getting top 4 or winner compared to the total that they are placing.

 

[Ban Heim]

Haven't read all of this, but shouldn't you take into account how many tournaments each character enters? For example, Meta Knight is used very often, so he'll score high because he's bound to place in the top 8 more than a character that isn't used often, like Samus for example.

There's got to be a way to factor this into the equation. That would require knowing every single character that enters each tournament though.

EDIT: Haha, just read some of the posts and this has already been addressed.

 

[Crizthakidd]

this seems pretty accurate with the most acepted tier list so far

 

[Facet]

Whoever said this statistical analysis is like a popularity contest is correct, however, its fairly obvious that the reason for a particular character to be popular would be that its very good. Therefore, I dont see a significant problem in that regard. Still, perhaps a more accurate title for this thread would be "List of most successful characters to date" or something.

 

[Lawlb0t]

Sure this is bias. But it should start to properly balance itself out with more data. Give it time. This isn't going to be 100% accurate, but its a good speculation on who's better.

Its the same thing with Melee. Luigi was high for a bit a believe, and then he flopped down after a year or two.

 

[Ark22]

This list shows some interesting trends in brawl. I wonder how character rankings will change as the game progresses.

It really surprised me how DK has placed better more often than Diddy Kong. Do people really dislike Diddy that much?

 

[henrytran]

THis rankings list is interesting when comparing it to the match-up chart in the tactical forum.

There's going to be flaws and biases in the rankings, but this list is what it is. Good job OP for taking the effort to put this together.

What would be really interesting is if you were able to keep this in some kind of database, so we could see how the list evolves over time. Though this is looking really long term, imagine something like this:

Spring 2008, Summer 2008, Fall 2008, Winter 2008, Spring 2009, Summer 2009, Fall 2009, Winter 2009, etc. Or maybe an entire year? Half year?

It would be interesting if some underrepresented characters begin to win more tournaments as the metagame develops further; by having only relevant results from a particular span of time, we ignore past results to see how which characters are doing well at any point in time. For example, we see Link at the very bottom right now. But let's say over the next year, he wins a good amount of tournaments. Due to the long-running list, he may still rank pretty low, but in the next year's within-itself comparison, he may end up middle or top.

 

[Kyari]

You know what I think you should do? Not to be bossy or negative in any way, because I really appreciate this list, but... if you could take your data (especially the final tier list) and organize it in an excel file, that would be top tier of you. Then you could so easily compare characters and make tier lists based on more than just pointage, but also how many tourney wins they got, how many top 4, etc. Not absolutely necessary, but I think it would be interesting to muse over the differences.

EDIT: Also YES, in reference to the above poster, (if you are familiar with Excel) it would be AWESOME if you could have a spreadsheet tab for each yearly quarter with respect to the final tier list. That would rock.

EDIT again: Xiivi? Fishers IN? How often are you there? That's only like half an hour away if i take 465 N.

 

[Xiivi]

I agree with Kyari's idea, it would be a very nice thing to do.
Also, providing links to the results of all of those tourneys would be great, I've wanted to look at some but for whatever reason I can't seem to find 2 or 3 of them. That would be very helpful.

 

[Zankoku]

Blah, I don't even know where half of the tourneys are any more. Anyway, I'm actually doing the calculations on an OpenOffice spreadsheet. I suppose I can add in rankings by tournament wins as well : P

 

[Kyari]

OpenOffice is alright, I think you should get Excel just so that everyone else can use your sheet because I'm sure more people pirate it than use another software, and as far as I know they don't go together (or at least not well...?)

Either way, spreadsheet is good, and if it isn't too much to ask, lemme see? :D

 

[Zankoku]

I PM'd it to you. Enjoy.

 

[feardragon64]

I can't take credit for the following post. I'm just posting for him because he came up with an idea that(if it's not too much trouble), would help balance the issue when people don't enter with a character as often. Anyways, here's what he said(lol I fully support this idea if you have time time to =]:

As a lot of people have mentioned, the current rankings are helpful but don’t take into account how many people have entered with a given character. For example, if a tournament has one Roy (I’ll use non-Brawl characters to avoid mucking things up with the real numbers) in the top four and that’s the only Roy entered, it means a lot more than if the same tournament has two Mewtwos in the top four and three-fourths of the pool is mewtwos.

Now, since its still early, that variation will distort statistics a lot. Later on, however, the disparity in how much a character is played will reflect rankings rather than distort them. That means the current system is important – even now, there very well could be a reason for the disparities that should be reflected in the rankings. However, providing the data both with the variation AND compensating for it would be pretty valuable.

To that end, might I suggest a supplemental ranking system?

For each tournament, take out the score per character – top 8s, top 4s, win – then divide it by the amount of entries with that character and multiply by the amount of total entries (which is equivalent to dividing by the fraction of competitors who were playing that character).

Basically,
(top8s + top4s*3 + win*5)/(charEntries/totalEntries)
unless I managed to screw up the weighting of the various positions, but that’s easily enough adjusted.

You up for it? =]

Edit: He is also forcing me to ask you this. If you decide to use this formula, can you call it the WARGLFARGL formula...?(I shouldn't have shown my friend the thread on tripping >>)

 

[Zankoku]

Impossible, or I'd immediately jump on that formula. Most tournaments don't pay attention to character mains past the top eight, and I can't just make up who mained what and hope I got it right.

 

[feardragon64]

Impossible, or I'd immediately jump on that formula. Most tournaments don't pay attention to character mains past the top eight, and I can't just make up who mained what and hope I got it right.

Ah, very sorry. My friend and I like sticking to our mains more often than not so I guess we kinda forgot about that

Well, thanks anyways for the consideration =]

By the way, is there a specific minimum tournament size that you're using? Surely a tournament with 32 people could allow for greater distortion(though I would imagine it wouldn't be a very big one) than, say for one with 64. Just a thought, though I can't really imagine any solutions past only picking large scale, trusted tournaments, which you're probably already doing. This is more me being curious about the standards for tournaments you're using. =]

 

[Zankoku]

Tournaments with good players will be selected, with a size of at least 32 so that we do get a reasonable amount of character representation all over the board, or decently well enough.

 

[Xiivi]

I like how you made separate lists in the first post excluding top 8 and top8/top4. That's very helpful as well. Anyway I could get that openoffice spreadsheet as well?

 

[Kyari]

He said he was having hosting bandwidth issues, so just use my webserver, I put his excel sheet on there.

http://68.58.87.34:8080/files/rankings.xls

that should work...for now.

 

[KernelColonel]

The idea of the tier list is character popularity, character tournament placement, and character playability.

Not many people play Yoshi. Metaknight is a good character. Metaknight wins many tournaments.

Therefore Yoshi is lower on the tier list than Metaknight.

Much lower, in fact.

This is actually as close to a tier list as we'll get, at least, until the BRoomers concoct the list.


I'm happy with how the list turned out. The gap is a little uneasing, but it's still wonderful to see G&W and ROB and Wario place so high (and to see that dreaded Pit in the C ranks). If there's one thing that I'm not happy to see it is Sheik's placement, what I mean is, it surprises me that less people are playing her (let alone winning).


LOL Ganondorf for Failtier.

 

[henrytran]

One thing I was interested in was number of appearances in top 8. Yeah, they aren't in top 4 or won, but they still in "the running" in terms of being close enough to maybe take a top spot.

So what I did with the excel was I added (# of appearances in top 8) + (# of appearances in top 4) + (# of appearances in top1).

What this gives us is "number of appearances in contender status" or whatever. Numerically speaking, we are putting less weight on winning and top 4, but still accounting for it. Below is the list.

Meta Knight 43
Snake 35
Mr. Game & Watch 21
ROB 19
Marth 17
King Dedede 16
Wolf 14
Pikachu 13
Ike 13
Olimar 13
Wario 11
Fox 11
Peach 9
Toon Link 9
Falco 9
Kirby 8
Lucario 8
Donkey Kong 8
Sonic 7
Ice Climbers 7
Pit 7
Diddy Kong 7
Lucas 7
Bowser 4
Zero Suit Samus 4
Jigglypuff 4
Mario 3
Ness 3
Ganondorf 3
Luigi 2
Captain Falcon 2
Sheik 2
Link 2
Samus 2
Yoshi 1
Zelda 1
Pokémon Trainer 1

At a glance, what we see is that Ike, Wolf, and Olimar higher up. ZSS a lot higher up. Kirbs and Lucario a lot lower. And MK beating Snake. Amongst other things.

 

[Tyr_03]

Very interesting that Wario is actually above Metaknight in tournament wins only. I wouldn't have expected that. Great data.

 

[Kyari]

One thing I was interested in was number of appearances in top 8. Yeah, they aren't in top 4 or won, but they still in "the running" in terms of being close enough to maybe take a top spot.

So what I did with the excel was I added (# of appearances in top 8) + (# of appearances in top 4) + (# of appearances in top1).

What this gives us is "number of appearances in contender status" or whatever. Numerically speaking, we are putting less weight on winning and top 4, but still accounting for it. Below is the list.

Meta Knight 43
Snake 35
Mr. Game & Watch 21
ROB 19
Marth 17
King Dedede 16
Wolf 14
Pikachu 13
Ike 13
Olimar 13
Wario 11
Fox 11
Peach 9
Toon Link 9
Falco 9
Kirby 8
Lucario 8
Donkey Kong 8
Sonic 7
Ice Climbers 7
Pit 7
Diddy Kong 7
Lucas 7
Bowser 4
Zero Suit Samus 4
Jigglypuff 4
Mario 3
Ness 3
Ganondorf 3
Luigi 2
Captain Falcon 2
Sheik 2
Link 2
Samus 2
Yoshi 1
Zelda 1
Pokémon Trainer 1

At a glance, what we see is that Ike, Wolf, and Olimar higher up. ZSS a lot higher up. Kirbs and Lucario a lot lower. And MK beating Snake. Amongst other things.

That's interesting too, so it's basically pure frequency in the top 8 then. Looks a little more accurate at the top, but downward past like 10th or so is still anyone's guess really, and probably still affected by popularity. I like this method of counting too, Ankoku. You could add a column for this fairly easily in your spreadsheet.

 

[Academic Sake]

One thing I was interested in was number of appearances in top 8. Yeah, they aren't in top 4 or won, but they still in "the running" in terms of being close enough to maybe take a top spot.

So what I did with the excel was I added (# of appearances in top 8) + (# of appearances in top 4) + (# of appearances in top1).

What this gives us is "number of appearances in contender status" or whatever. Numerically speaking, we are putting less weight on winning and top 4, but still accounting for it. Below is the list.

Meta Knight 43
Snake 35
Mr. Game & Watch 21
ROB 19
Marth 17
King Dedede 16
Wolf 14
Pikachu 13
Ike 13
Olimar 13
Wario 11
Fox 11
Peach 9
Toon Link 9
Falco 9
Kirby 8
Lucario 8
Donkey Kong 8
Sonic 7
Ice Climbers 7
Pit 7
Diddy Kong 7
Lucas 7
Bowser 4
Zero Suit Samus 4
Jigglypuff 4
Mario 3
Ness 3
Ganondorf 3
Luigi 2
Captain Falcon 2
Sheik 2
Link 2
Samus 2
Yoshi 1
Zelda 1
Pokémon Trainer 1

At a glance, what we see is that Ike, Wolf, and Olimar higher up. ZSS a lot higher up. Kirbs and Lucario a lot lower. And MK beating Snake. Amongst other things.

You should weight this so that the differences between characters would be more distinct. For example if you make it in the top 8 you give that character a weight of 1. If they make it into the top 4 you weight it with 3 and top 1 gets a 7. Or however. This way it'll account for both frequency and severity.

 

[Kyari]

@ Academic Snake: That's what the original topic creator did.